Adaptor House English Dictionary
AHED
Authored lexicography, built on first-order logic.
AHED is the declared end goal of the dictionary work and the largest record producer this ecosystem will meet — headword scale, not line scale. It is not a preservation corpus and not a word list. Its claim is that a dictionary entry can be an assertion with refutation conditions rather than a description, and that the terms the dictionary uses to describe itself must meet that standard first.
- Status
- Registered — no entry
- Publication
- Not published; no edition tagged
- Id
- Space declared; minting blocked
- Verification
- Pending
Each empty slot above has a reason, and the reason is part of the record. Nothing here is withheld; where a value is absent, the absence is stated.
Id. AHED has a declared id space. The prefix AHED-
is reserved so that nothing else reaches for another entity's ledger to name an English
dictionary entry — the ordering that matters most at headword scale, and the reason the
space was declared before the first entry rather than after. What is not settled
is the shape after the prefix, and minting is blocked until it is.
Why the grammar is open. A conventional dictionary identifier would be built from headword, homograph and part of speech — and part of speech would in turn presuppose categories like noun, verb, and adjective. Every one of those is a borrowed term, carried in on familiarity rather than derivation: the part-of-speech categories are historical English's own grammar, a classification scheme this system did not derive and has no standing reason to inherit intact. None of them — not headword, not homograph, not the part-of-speech categories beneath it — holds a derivation here. AHED is not a conventional dictionary rebuilt on logic underneath the same categories; whether anything resembling those categories survives is exactly what is undetermined. Choosing an identifier shape assembled from vocabulary nobody derived would make the identifier look decided while denoting nothing. So the open work is deriving the metalanguage from first principles, and the shape — and whether a unit answering to "headword" exists at all — follows from what those derivations turn out to support, not from what traditional lexicography assumes.
Not historical English, formalized. Architect direction, 2026-08-15: AHED will not have nouns, verbs, adjectives, and the rest of the part-of-speech inventory, because those categories are themselves part of a historical English long outdated — inherited grammar-book vocabulary, not derived terms. The gate this page names is not "define headword rigorously and proceed as before." It is open whether AHED ends up organized around anything nameable as a headword, and open independently whether any surviving unit is disambiguated by anything resembling part of speech. A page that presented either as settled would misrepresent the ruling it is meant to reflect — which is precisely the defect the archived prior version of this page exhibited, preserved at Historical Present for the record rather than silently overwritten.
What the dictionary is for
A conventional dictionary entry reports usage. It is answerable to a corpus in the loose sense that a lexicographer read one, but an individual sense line is rarely something a reader can test. AHED's position is that a definition should be an assertion with stated refutation conditions — something a corpus can contradict.
That is why the dictionary is built on first-order logic rather than decorated with it. The logic is not a display layer over prose written first; it is what the entry is, and the prose continues it.
A name bound by evocative fit cannot be refuted; a name defined as the evaluation of a logical string can be.
The same standard applies reflexively, and that is the load-bearing part. The terms AHED uses to structure itself are subject to the rule before its contents are. A dictionary that derived its entries rigorously while describing them in borrowed vocabulary would be sound in appearance only.
Soundness
Conformance is checked through the Trope Compendium's existing apparatus rather than a second one built for the dictionary. Its predicate vocabulary is a registration list, not a pattern: a term is admissible once it has been registered from a corpus, and a string containing an unregistered term is rejected rather than passing quietly.
That is the mechanism behind the claim above. It is also precisely why headword cannot simply be used — an unregistered term fails the check, which is the check working.
Boundary with Language Preservation
Adaptor House also runs language-preservation work that produces English lexicon records. Those are different objects, and the boundary is written down rather than left to be inferred: preservation records are consent-bound elicitation, governed by a field protocol and a consent record; AHED entries are authored lexicography for publication, carrying editorial obligations only.
Neither space mints into the other. An AHED entry may cite a preservation record as a source, and citation is not identity. The boundary was declared while neither space had minted anything — the only moment the choice was free.
What exists today
Stated plainly, because a registered surface that implied more would be the failure this project exists to name:
- A registered project node and a declared id space, with a reserved prefix.
- A minting path that refuses, and names the open decision and its owner when it does.
- A standards file that is the single authority on the grammar, currently recording that grammar as undetermined.
- No organizing unit named — headword or otherwise. No entry. No edition.
Coinage of whatever unit the derivation licenses — a headword-like object, or something historical English never named — belongs to the author of record. It is not a session's to pick, and it is not the first gate in any case — the metalanguage is, and that metalanguage is not obliged to reproduce headword, homograph, or part of speech merely because those are the terms a reader would expect.